Worker distribution strategies in PyTest - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of pytest-xdist with loadscope distribution

Given the following pytest test file and running pytest with pytest -n 2 --dist=loadscope, what will be the order of test execution across workers?

PyTest

import pytest

@pytest.mark.scope('module1')
def test_a():
    pass

@pytest.mark.scope('module1')
def test_b():
    pass

@pytest.mark.scope('module2')
def test_c():
    pass

@pytest.mark.scope('module2')
def test_d():
    pass

A Tests are randomly distributed without scope grouping

B Worker 1 runs test_a and test_b; Worker 2 runs test_c and test_d

C Worker 1 runs test_a and test_d; Worker 2 runs test_b and test_c

D Worker 1 runs test_a and test_c; Worker 2 runs test_b and test_d

💡 Hint

Loadscope groups tests by their scope marker to minimize setup overhead.

The loadscope distribution groups tests by their scope marker (here 'module1' and 'module2'), so tests with the same scope run on the same worker.

❓ assertion

intermediate

Correct assertion for balanced worker load

You want to assert that pytest-xdist distributed 10 tests evenly across 2 workers. Which assertion correctly verifies this?

A assert worker_test_counts[0] == 10 and worker_test_counts[1] == 0

B assert all(worker_load == 5 for worker_load in worker_test_counts)

C assert sum(worker_test_counts) == 10 and max(worker_test_counts) - min(worker_test_counts) <= 1

D assert min(worker_test_counts) == 0

💡 Hint

Balanced load means tests are nearly evenly split, allowing a difference of one.

Option C checks total tests are 10 and the difference between max and min tests per worker is at most 1, which is correct for balanced distribution.

🔧 Debug

advanced

Identify the cause of uneven test distribution

A test suite with 12 tests is run with pytest-xdist using --dist=loadfile and 3 workers. The distribution is uneven: one worker runs 8 tests, others run 2 each. What is the most likely cause?

A The number of workers is greater than the number of test files

B pytest-xdist randomly assigns tests ignoring files

C Tests are grouped by markers, causing imbalance

D Tests are grouped by file, and one file contains 8 tests while others have 2 each

💡 Hint

Loadfile groups tests by their file location.

With --dist=loadfile, pytest-xdist assigns all tests from the same file to the same worker, so if one file has many tests, that worker gets more load.

❓ framework

advanced

Choosing the right pytest-xdist distribution mode

Which pytest-xdist distribution mode is best when tests have expensive setup per module and you want to minimize setup duplication?

A loadscope

B loadgroup

C no distribution (run sequentially)

D loadfile

💡 Hint

Consider grouping tests by their setup scope.

loadscope groups tests by their scope (like module), so setup runs once per worker, minimizing duplication.

🧠 Conceptual

expert

Impact of worker distribution on flaky tests

How can improper worker distribution in pytest-xdist contribute to flaky test results?

A Tests sharing state run on different workers causing race conditions

B Tests run sequentially causing timeout errors

C Workers share the same memory space causing data corruption

D Tests are always isolated, so distribution has no impact on flakiness

💡 Hint

Think about shared resources and parallel execution.

If tests that share state run on different workers without proper isolation, race conditions can cause flaky failures.